| With the rapid development of global economy and trade,shipping industry is playing an increasingly important role in international trade and transportation,and the emission pollution caused by engines has been widely concerned.In order to meet the current increasingly stringent emission regulations,dual-fuel marine engines fueled by natural gas and diesel are favored by researchers.However,the two-stroke pre-mixed dual-fuel marine engine is prone to knocking at full load or poor cooling conditions in gas mode,which not only affects the overall dynamic and economic performance of the engine,but also limits the increase of compression ratio and affect the further increase of engine power.Therefore,this thesis takes Win GD7X82 DF low-pressure gas injection dual-fuel marine engine as the research object.Through numerical simulation technology,aiming to investigate the effect of different engine operating parameters on engine dynamic,economic performance and combustion performance.Response surface methodology and two multi-objective optimization algorithms were used to optimize parameter settings to improve engine combustion and economic dynamic performance.In this thesis,a one-dimensional simulation model of the 7X82 DF dual-fuel engine was built using simulation software AVL-BOOST,and the results showed that the errors of the main parameters were all less than 3%.The one-dimensional model with 100% load of natural gas model was selected to conduct simulation research,exploring the effects of engine operating parameters start of combustion timing(SOC),gas injection pressure,mass of diesel and in-cylinder water injection technology on engine economic,power and combustion performance.Selected natural gas injection pressure,SOC and fuel quality as independent variable parameters,engine power,fuel consumption,peak firing pressure and ringing intensity as target parameters,and 125 groups of one-dimensional simulation data of independent variable parameters as database.And taken response surface analysis method to establish the response surface prediction model of each target parameters.The objective of optimization was to reduce the knock tendency of the engine as far as possible without increasing fuel consumption,and we took ringing intensity(RI)as the knock intensity evaluation index.The optimization results showed that when the SOC was-7.61°CA ATDC,the natural gas injection pressure was20.0bar,the mass of diesel was 14.03 g,the corresponding brake-specific fuel consumption was156.443g/k W × h,which was reduced by 3.46%,the power was 22434.8k W,which was reduced by 4.27%,and the RI was 4.1779,the knock intensity decreased by 11.90%.The knocking tendency of engine was greatly reduced under the condition of sacrificing little engine’s power.To solve the power decline problem after response surface optimization,the established target response surface prediction model was used as the objective function replacement model,combined with the multi-objective particle swarm optimization algorithm and multi-objective gray wolf algorithm to optimize the operating parameters.And with the ringing intensity as the limit,to obtain as high power and as low as possible fuel consumption and ringing intensity value as the goal.Compared with the optimization results,the power of 22668.0k W was obtained by multi-objective particle swarm optimization,which increased by 0.61%,the brake-specific fuel consumption was 156.256 g/k W h,which decreased by 3.58%,the ringing intensity was4.4326MW/m2,and the knocking tendency was reduced by 6.49%.After multi-objective grey wolf algorithm optimization,the obtained power is 22691.7k W,which increases by 0.713%,the obtained brake-specific fuel consumption is 156.158g/k W h,which decreases by 3.67%,and the ring intensity is 4.4118MW/m2,which decreases by 6.92%.This indicates that the multi-objective grey wolf algorithm has better optimization ability.After multi-objective grey wolf algorithm optimization,the economic power of the engine is improved,and the engine knocking tendency is further reduced. |